Native DSD files, HQPlayer & Mytek Stereo192-DSD

[manisandher]

"However I find the NAA approach works very well. For files that are played without any changes its pretty much indistinguishable from playing the same file from a NAS with mpd on the same box as NAA."

 

Thanks Demian, that's reassuring to know.

 

But could I place you on the spot? I've never heard a DAC being fed by the Mykerinos card. Could the NAA approach be still outdone by the Mykerinos card in a computer with Windows?

 

Mani.

 

[Miska]

Hopefully something just needs bedding in because I totally love the sound I'm getting, but just wish it were a bit sweeter and more laid back.

 

Mytek is a magnifying glass for all the detail, good or bad. My first impressions (without NAA) was that it doesn't have a sound of it's own at all. Just extreme detail in a neutral way.

 

Burn-in will tame some of it's initial hardness...

 

The sound I'm getting through the NAA (with PCM or DSD) is almost 'too clean' and dynamic! I find it really hard to describe in any other way. It's not 'bright', or 'forward' or 'edgy'. Rather it's so clean and dynamic that it's almost tiring to listen to. Not relaxing at all. But it certainly keeps your attention at least.

 

That was my initial impression too when it arrived here, again without NAA. I've heard others describe it this way too.

 

[Miska]

I've never heard a DAC being fed by the Mykerinos card. Could the NAA approach be still outdone by the Mykerinos card in a computer with Windows?

 

Would it be much different from a computer equipped with RME or similar interface? What comes out of any card placed inside computer depends to some extent on that computer.

 

I'm taking the other extreme as an experiment, I'm putting NAA inside my DAC and feed I2S straight from there... (luckily I have some vertical space inside the box, since it's a 19" 2U rack box, because the PSU caps didn't fit in 1U)

 

 

[ted_b]

a newly minted Mytek DAC. It has so much potential (due to the detail and dynamics that are evident) but you just wish it had more musicality or richness in the midrange. At several hundred hours it "loosens up" very nicely. I would think that if the NAA acts as an even better scrubber, then the pristine nature of the Mytek might be a bit much on day one or two. But for analog-DSD jazz reissues (KOB was your first impression) the analog warmth is there and comes through like a breath of fresh jazz-tinged small club air (meaning slightly illegal in most states, except medicinally).

 

[elcorso]

I can't afford this! I'm on ELS speakers (Martin Logan CLX), it can kill my ears.

 

I prefer some muddiness in the music, like the analog way, then never too polite.

 

Ted recommendation SACD reissues from analog master tapes, like Bill Evans Trio: Waltz for Debby, CAPJ 9399SA (Analogue Productions SACD) is a good test.

 

Another good test is a solo violin recorded in DXD, transferred to DSD in this very good SACD (available in Amazon UK):

 

http://www.amazon.co.uk/Raasted-Violin-Sonatas-Solo/dp/B0040Y7ERO/ref=sr_1_1?s=music&ie=UTF8&qid=1328218416&sr=1-1

 

Also in Amazon US:

 

http://www.amazon.com/Solo-Violin-Sonatas-Raasted/dp/B0040Y7ERO/ref=sr_1_1?s=music&ie=UTF8&qid=1328220541&sr=1-1

 

Ted, regarding "...(meaning slightly illegal in most states, except medicinally)....". In my country they are studying to legalize it, trying to stop the traffic to the US & EU from Colo***a, since we don't have any army and low budget for policeman.

 

Roch

 

[manisandher]

"I would think that if the NAA acts as an even better scrubber, then the pristine nature of the Mytek might be a bit much on day one or two."

 

Yes, I'm sure you're right about the NAA perhaps exacerbating the Mytek's pre-burned-in nature.

 

"Some claim that a NAS won't be as good as a local drive..."

 

Haha, this is the culprit! I've just copied my DSD folder to the local drive where HQPlayer sits. And low and behold, the 'clean and dynamic' but ultimately unlistenable sound has gone.

 

I have an 8-port Gigabit switch. I can't see how 5.6448Mbps of data from the NAS to the playing PC, and then the additional 5.6448Mbps of data from the playing PC to the NAA could 'bother' the switch. There are certainly no drop-outs or anything, but maybe a higher load on the switch increases noise entering the NAA via the LAN? Don't know. But I'm now inclined to go for a better switch, with maybe optical isolation. Any recommendations for a combined copper/fiber-optic Gigabit switch?

 

Mani.

 

[manisandher]

Thanks Roch. I've gone ahead and ordered the Raasted SACD. I would order the Waltz for Debby too, but I've listened to this way too often in the last few months - need a bit of a break from it.

 

I'll look into some of the other recommendations that both you and Ted have put forward.

 

Cheers,

Mani.

 

[losingmyreligion]

 

"Are you suggesting that a 3 box approach may be sonically superior to your yet to be released L1000?"

 

AZ,

I can't (yet) speak to Demian's L1000, but here's how I look at it.

 

For those that need / want a high level of processing, such as users of HQ, iTunes, XXHE(?), the 3 box approach is "a" way of isolating the noise caused by the processors.

 

If you're okay with having an iPad, laptop, iphone (or similar) drive an mpd client (for use with the PK90), you may hit diminishing returns in going with the 3-box approach.

 

 

I'd love to hear a plain vanilla iTunes user interface feeding (via wired ethernet) a tiny linux appliance, which processes the data and sends it out via Firewire (and Async USB, I suppose) to the DAC.

 

Not sure if Demian's L1000 will allow more horsepower at a similar sonic performance, or whether it will further lower the noise footprint as compared to the PK90. Demian?

 

 

clay

 

 

 

 

 

[losingmyreligion]

"Some claim that a NAS won't be as good as a local drive..."

 

I don't see how a NAS could sound better than a properly setup local drive. OTOH, I've never bothered with a NAS, due to that belief. Maybe optical isolation of a NAS could be better?

 

Was thinking of getting a Synology to use for backup however.

 

clay

 

[Miska]

Any recommendations for a combined copper/fiber-optic Gigabit switch?

 

I've mostly used HP and 3Com rack-mount switches with fiber (too expensive and acoustically noisy for home use, IMO). But D-Link has fairly inexpensive managed switches with QoS and SFP-slot. As well as SFP modules.

 

Another option are media converters, I've used those in the past for measurement systems in industrial environment, things like this:

http://www.moxa.com/product/IMC-101G.htm

They have also switches and SFP-modules:

http://www.moxa.com/product/SFP-1G_Series.htm

(1000Base-SX using LC-MM fiber is the most typical)

 

 

[elcorso]

From 2008 analog master tape on Studer A80 to EMM Labs by Pyramix Virtual Studio. From PentaTone Classics (PeterSt N.L.)

 

Amazon UK:

 

http://www.amazon.co.uk/Beethoven-Piano-Sonatas-Nos-21/dp/B0000AJF9Y/ref=sr_1_2?s=music&ie=UTF8&qid=1328238805&sr=1-2

 

Amazon US:

 

http://www.amazon.com/Beethoven-Piano-Sonatas-Nos-Hybrid/dp/B0000AJF9Y/ref=sr_1_1?s=music&ie=UTF8&qid=1328239385&sr=1-1

 

At at the 'nice price' yet...

 

Roch

 

 

[losingmyreligion]

 

 

"- noise and other pollution in the music server don't matter"

 

you could still pollute the rest of your system (and therefore it's sound) via the AC circuit, and perhaps via airborne EMI/RFI if the boxes were in very close proximity, IMO.

 

clay

 

 

 

[ted_b]

Pentatone SACD's. I own 4 of 7 of them and Pentatone has them all listed (Beethoven piano sonatas except for her Chopin concerto with husband Kent Nagano) as "DSD RECORDED". SA-CD.net concurs that these are pure DSD recorded.

 

 

Regardless, they are some of my main go-to DSD files. They were the piano tracks that made me first take notice of DSD's amazing realism. I love them all.

 

[RealAudio]

The enterprise NAS boxes (netapp by far the market leader) are commonly connected of sometimes multiple 10 gb link. In addition, they have very large very smart redundent caches. The end result is they are typically faster than local disks. Home units aren't nearly as sophisticated, so your mileage may vary.

 

[1audio]

Mani asked: " Could the NAA approach be still outdone by the Mykerinos card in a computer with Windows?" . Reducing that to an a-b comparison is very difficult at best. There are so many variables to remove before you are actually answering the question.

 

The Mykerinos and the Lynx and the RME cards were designed for production, not for listening. They have very good capabilities and the flexibility to work with Pro Audio equipment but the core value is 8 or 16 or 32 or 64 inputs, in some cases with sample rate conversion on each input. They allow connection to devices like the Pacific Microsonics with its odd ball Dual Link interface and can provide a very good signal to those DAC's. They have no special magic that will make the signals "better". They do support remote sync which can help the jitter issue but many of today's DAC's have really good jitter reduction already.

 

The recommended system for Mykerinos + Pyrimix is a specific Intel quad core with a specific version of Windows, hacked to give the "dsp" component exclusive access to 2-3 of the 4 cores. (This may have changed.) Comparing that to an ultralight Linux install (Alix?) to the same DAC seems fundamentally odd.

 

But continuing on this thought experiment the next question is if the content needs any processing or is just getting passed to the dac. I don't believe there are drivers for the Mykineros on Linux so we have the core difference between OS and card and support system. With the ultralight system + NAA we can make it a very small emi/rfi/conducted radiation source (where the L1000 lives). If your DAC + analog chain are insensitive to the radiation then that part won't matter. (I don't see that happening.) So do you encase the Win + MYK package in a really good shield enclosure (a major project especially with the heat output)? And if you do and keep the processing minimal then the remaining comparison would be the card. But to get the data in or out of that shield enclosure (AES? Ethernet etc.) will affect the data quality most likely.

 

Finally you have the issue of a big PC box in the space next to the DAC. If the DAC is in your listening room then this box is there as well. Its physical presence will affect the sound in the room.

 

I would say that moving the processor box out and using an NAA + lightweight Linux box would sound better just because it can have a smaller acoustic impact on your space. I really can't say how the other issues would work out. Too many possibilities. Its almost a Semi + trailer VS. Lotus 7 to the supermarket. Neither are the right vehicle. The 7 will force you to leave some stuff behind (no room) and the Semi is hard to park to say the least. I know which I would take however.

 

 

 

[1audio]

Sometimes the big things actually distract from the problem. I have not seen an issue with NAS's performance playing audio. The data arrives fast enough that it should not be an issue with even a cheap NAS. But as Miska points out configuration of the system can be a big problem. So can the noise from network cables etc. Or perhaps the NAS is too close and its radiating a lot of noise. Meeting FCC does not mean a device is noise free. It just means that the nearby TV stations will get through. Bigger faster NAS's are intrinsically noiser.

 

Some people like the effect of the noise. They may not know it but they like it.

 

I'm not convinced that going to optical network links will really fix anything since the devices at each end will still radiate. Sometimes more than the original would have. And you need power for the optical receiver if its outboard (and a much bigger power burden inside if its internal).

 

The NAA has about a 2 second buffer so if the data is intact then any affect upstream of the NAA would be affecting sounds 2 seconds before the processing actually happens. I.E. the last 2 seconds would sound better. Perhaps Miska can adjust the buffer to be much bigger so its a form of memory player? The response time to control input would suck big time (or he will need to redesign the system). If its affecting the data then Miska would really want to know.

 

[PeterSt]

Finally you have the issue of a big PC box in the space next to the DAC. If the DAC is in your listening room then this box is there as well. Its physical presence will affect the sound in the room.

 

This looked like a side note, but it is of major importance.

 

I just wanted to emphasize this.

Peter

 

[PeterSt]

I'm writing this on the same desktop PC and no matter what I do, I can't detect any affect on the sound whatsoever.

 

Which at least for me implied that all influences had gone. Not that you said this Mani, but you really wanted to ...

 

I guess that by now it is fair to say that this is not true.

 

Back to square one ? No, that sounds crude. But it all just doesn't work like this. Possibly the base is now better, but what has to be done still has to be done.

Talking in secrets ? mwah, but all I want to say is that you can't solve these issues by means of hardware only.

I think if it could, I had done it myself by now. And Mani, think about the reason why XXHE shuts off the network and/or why it is do damn difficult to use a decent remote because it needs to be outside something like RDC ...

 

Peter

 

[PeterSt]

But I have a problem. The sound I'm getting through the NAA (with PCM or DSD) is almost 'too clean' and dynamic! I find it really hard to describe in any other way. It's not 'bright', or 'forward' or 'edgy'. Rather it's so clean and dynamic that it's almost tiring to listen to. Not relaxing at all. But it certainly keeps your attention at least.

 

I know the status of this changed throughout the thread, but honestly, doesn't this look like WASAPI sound (though XXHE) ? I think it does.

There is no way you can point out it being wrong, but still it is.

I told it before, the first XXHE WASAPI version (which never went to the public) in the beginning of 2007 was so perceivedbly accurate that instruments fell apart. This is a known phenomenon in audio. Can we say it is too good (accurate) in such a case ? I don't think so. Make it more accurate and it disappears.

 

My worry now is that what I'm getting from the NAA are false dynamics.

 

Us knows us, and for others it may be difficult to get the real merits of such an expression, but in communication with Mani (just by this sentence) I know enough. When there's this worry, you better think in advance it's justified.

 

Haha, this is the culprit! I've just copied my DSD folder to the local drive where HQPlayer sits. And low and behold, the 'clean and dynamic' but ultimately unlistenable sound has gone.

 

Ok, apologies I'm now talking like on my own forum etc., but ...

So now we are there ? now *that* would be a coincidence. So no, you will merely have reached the other side of the ultimate and day after tomorrow things are dull again or not accurate enough. Anything.

Please keep in mind : only rely on statistics and the coincidence will not exist that this now is good. Merely, there is no reason that this now suddenly is good. All is against it. So it will be wrong. Just watch that day (I always take 5 at least as you know).

 

If it's not the Mytek, it must be the NAA.

 

I never heard the thing of course, but I do agree. Us knows us again.

So, no matter what others are constructively talking you into, it won't be true and which by now has been proven already. Oh, the Mytek will improve alright (ok, I suppose), but we have other means to judge these merits. This too is a matter of getting used to or skilled in. I just tried to do it by reading, FWIW ...

Btw, the last thing I'd like to confirm is "that it's the NAA". It's only that it doesn't solve all, as how it seems.

 

Outside the above letting myself go a little (just for fun because it's totally useless of course), we now must wonder how to get this right. Ehm, no, how to KNOW WHEN it is right.

I don't think many possibilities for this exist. Well, don't encounter ANY anomalies, which was my first question in this thread. And again, is this DSD ? most probably not, or not yet anyway. And then to think I was hoping for a definite answer on this. Well, that has to be postponed a little, I'm afraid.

 

Peter

 

[Miska]

I'm not convinced that going to optical network links will really fix anything since the devices at each end will still radiate. Sometimes more than the original would have. And you need power for the optical receiver if its outboard (and a much bigger power burden inside if its internal).

 

Industrial systems by "standard" use 24 VDC power supplies. And I've been dealing with several occasions where you really need to isolate both end points, not least because heavy electrical machinery can make AC feeds and even ground potentials swing +-20VDC... Also things like big electrical motors in places like trains can emit pretty nasty stuff into sensitive measurement systems.

 

Now, home is kind of cozy and friendly environment compared to those, but that's why it's possible to reach quite incredible resolutions.

 

NAA is kind of inverse of an audio analysis system I designed in late 90's where high sensitivity ADCs were networked (roughly 80-120 dB gain), partially for the same reasons.

 

The NAA has about a 2 second buffer so if the data is intact then any affect upstream of the NAA would be affecting sounds 2 seconds before the processing actually happens. I.E. the last 2 seconds would sound better. Perhaps Miska can adjust the buffer to be much bigger so its a form of memory player?

 

FIFO in NAA is now one second. This is about maximum tolerable lag. Yes, it can be adjusted but would make things like digital volume control and seeks completely unusable.

 

 

[PeterSt]

... I will try to point something out. Could be hard to get for others, but Mani should be able to digest it;

 

Recently I managed to improve on SQ again. Let's say that the freshness (as heard through the NOS1) increased by a factor of two. Ridiculously actually. And, all is still 100% OK to my ears, no matter the balance changes so much. It's just heading for the day that cymbals sound as loud and good as they do in real life (but when that really happens it will be too loud undoubtedly - just saying).

 

So, yesterday I was testing a few things, and the volume was still at the playback level from the night before. Nobody in the house, so I let it be. The PC was in Minimize OS mode, and the only things unreal were Mixed Memory Mode and Attended Playback (meaning the UI being operative).

 

It was totally completely unlistenable. I had to lower the volume or otherwise my ears would hurt.

 

Of course we already know what the various settings in XXHE do, and the both just mentioned for sure will be recognizeable by everybody using XXHE. Mani will too. But mind you, these are settings to again improve on SQ, which is different from creating the difference between "no anomalies" and "totally unlistenable". Well, since my recent improvement now suddenly this crazy distance is there. So ...

 

So I improved on something (and mind you, twice the freshness *is* crazy), but when not all is done to have it all (whatever that is) right, and it turns out against you. Really really against you.

The very similar happened when the first NOS1-USB had to go out. It suddenly needed the Minimize OS functionality or otherwise the thing sounded bad only. Not as bad as now with mentioned settings, and this is only because I improved on something.

 

So see ? It is not only that things are difficult again, but merely that you can't do without those "dials". Well, says me.

When this hadn't happened (with the NAA and stuff) I only could have said that only for a very few such things can help/work (referring to Miska). I'm not saying that by now this is not true either, but I do keep up that is will be for a very few only, if true at all. Remember what I said earlier in the thread : with no dials there is no way to know. But moving a few discs around is also a "dial" ...

 

2c

Peter

 

[Miska]

The PC was in Minimize OS mode, and the only things unreal were Mixed Memory Mode and Attended Playback (meaning the UI being operative).

 

I'm still wondering why you are keeping the audio playback together with everything else and the player. Why you don't for example send the entire file as PCM to the NOS1 and disconnect the wire? I could easily do that with NAA.

 

Noisy PC connected to the DAC? Do you even have galvanic isolation inside the NOS1?

 

you can't do without those "dials"

 

Dial WHAT?

 

 

[Miska]

think about the reason why XXHE shuts off the network

 

Why don't you shut down firewire and USB too? By your description XXHE would at it's best when it's running in powered-down computer. Well, each to their own, but would you please put your XXHE stories into some other thread.

 

I'm personally completely uninterested about any player that would shutdown my network, because I want to control music playback from the same computer while programming new version of HQPlayer and writing here at CA.

 

NAA just disconnects the player computer from the actual audio playback. At least it's better than "network streamers" or "music servers" because it doesn't have to perform any of the file reading/decoding/etc of those. And it doesn't need to have any HDD either.

 

I'm running it on a 500 mW ARM-based SoC too, fits into a DAC and runs from a 5VDC linear supply (the CPU itself runs from 1.6V line). It becomes just a DAC with Ethernet connection instead of Firewire or USB. Why would you use USB or Firewire when you can use Ethernet too and get galvanic isolation for free?

 

 

[PeterSt]

Well, each to their own, but would you please put your XXHE stories into some other thread.

 

I am sorry if you feel it like this Miska. This is not about XXHE stories at all, but in the end about your means of getting out the noise against mine. Well, it always has been, and now there's the subject. By coincidence.

 

What's also more difficult to see (if at all) is that not only you and not only me, but also some people like Mani hunt for the best. So ...

 

So, solutions like yours have been talked through more than once (Mani being the trigger of it in many cases), and now allow me to see the relation. No need to explicitly talk about it, only you don't know it. Ok, now you do. There's also no need to wonder why *he* out of all made the combination subject to this thread, right ?

 

Please read back into my first post / first sentence in this thread and what I tell there (about you). No need to be annoyed.

But I really hoped it worked out and still do (believe it or not) especially because I couldn't do it myself.

 

I hope all is right now ...

Apologies,

Peter

 

 

[Miska]

I personally prefer if threads stay on the subject.

 

If you want to start a player or DAC comparison thread, please start it with a suitable subject.

 

I prefer Mytek DAC thread to talk about Mytek DAC not about NOS1. And I prefer JRiver thread to talk about that and I don't go there bashing JRiver and touting about my software instead.

 

I just get annoyed if my mailbox gets flooded with messages of topic X where the content is a mile long posting about Y.

 

 

Thanks for understanding.